Long-Term Follow-up of Children for a 2-Year Period to Confirm the Safety and Immunogenicity of GSK 257049 Vaccine: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Long-Term Follow-up of Children for a 2-Year Period to Confirm the Safety and Immunogenicity of GSK 257049 Vaccine (NCT00323622), a randomized phase 2 study. The primary outcome is Number of Subjects With Serious Adverse Events (SAEs).

2. Clinical Question

This is a follow-up safety question, not a test of vaccine efficacy. The registry also lists antibody measurements, malaria episodes, time to malaria episodes, and anemia among the secondary outcomes, but the primary endpoint is the number of subjects with serious adverse events. The study description notes that the extension followed children from the earlier Mozambique study; it did not recruit new subjects or administer additional vaccine.

3. Trial Design

This completed phase II study was randomized and used a parallel-group arrangement, with no masking. It enrolled 1,737 participants. Its stated purpose was long-term follow-up of children for a two-year period to assess safety and immunogenicity of RTS,S/AS02A, with malaria efficacy also included among the follow-up objectives described in the registry.

The extension followed children already enrolled in NCT00197041. Children had received their assigned vaccines in that earlier study, with three doses given at Months 0, 1, and 2. The extension period began 21 months after Dose 1 and continued through Month 45. No new participants were recruited, and no vaccine was administered during the extension. Thus, the reported groups reflect prior study cohort and vaccine assignments rather than new treatment assignments made during this follow-up.

Groups are separated by cohort, vaccine history, and age at first vaccination. RTS,S/AS02A groups include children younger than 24 months and children aged 24 to 48 months at first vaccination. Comparator groups include younger children who received Prevnar and Hiberix, and older children who received Engerix-B. Two cohorts from the earlier study are represented, with Cohort 2 described as followed for malaria-disease analysis. The group sizes vary, and the available comparisons use the Cohort 1 RTS,S/AS02A group of children younger than 24 months as their reference.

The primary safety analysis covered the Total Vaccinated cohort: children vaccinated in NCT00197041 who re-enrolled for this follow-up and had data available. The registry record does not report the original sample-size justification, details of the randomization procedure, or a prespecified statistical analysis plan for the extension. The follow-up was open, so masking was absent.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Once re-enrolled, subjects in this study were allocated to the same groups as in the NCT00197041 study, as well as the same cohorts, e. a. Cohorts 1 and 2 whose subjects were followed for analysis of, respectively, malaria infection and disease.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis was performed on the Total Vaccinated cohort, which included all subjects vaccinated in the primary NCT00197041 study, and re- enrolled in this follow-up NCT 00323622 study, and for whom data were available.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Subjects With Serious Adverse Events (SAEs), measured over Throughout the entire study period: from Month 21 to Month 45 (Month 0 = administration of Dose 1 of RTS,S/AS02A or comparator vaccine in the NCT00197041 study)..

The primary endpoint is the number of subjects with serious adverse events during the entire follow-up period, from Month 21 to Month 45. Month 0 refers to the administration of Dose 1 of RTS,S/AS02A or a comparator vaccine in NCT00197041. The registry defines serious adverse events as medical occurrences that result in death, are life threatening, require hospitalization or prolong hospitalization, or result in disability or incapacity.

This endpoint counts participants who experienced a serious event, rather than counting every event episode. It is used to describe clinically consequential events over the follow-up period and to compare their observed frequency across the registered groups. The reported group results include both the number of participants with an event and the number at risk in each group.

The endpoint is not a comprehensive measure of every aspect of safety. It does not, by itself, describe the severity or duration of each event beyond the registry definition, establish whether an event was caused by a vaccine, or show the full pattern of non-serious events. The registry lists serious-event terms including Plasmodium falciparum infection, anemia, febrile convulsion, pneumonia, gastroenteritis, and acquired immunodeficiency syndrome. These terms provide context for events recorded as serious, but do not replace the primary participant-count endpoint.

Because this is an extension after the earlier vaccination period, its results describe serious events recorded during follow-up, not events arising from a new dose given in this study.

7. Results Posted in the Registry

• Cohort 1-RTS,S/AS02A ≥24M Group: 3.3% (17/515) versus 11.9% (21/176) on control; risk difference -8.6 percentage points (95% CI -13.7 to -3.6), risk ratio 0.28 (0.15–0.51); unadjusted two-proportion z-test p = <0.001.
• Cohort 2-RTS,S/AS02A <24M Group: 2.4% (1/42) versus 11.9% (21/176) on control; risk difference -9.6 percentage points (95% CI -16.2 to -2.9), risk ratio 0.20 (0.03–1.44); unadjusted two-proportion z-test p = 0.065.
• Cohort 2-RTS,S/AS02A ≥24M Group: 0.0% (0/134) versus 11.9% (21/176) on control; risk difference -11.9 percentage points (95% CI -16.7 to -7.1); unadjusted two-proportion z-test p = <0.001.
• Cohort 1-Prevnar-Hiberix <24M Group: 15.1% (24/159) versus 11.9% (21/176) on control; risk difference 3.2 percentage points (95% CI -4.2 to 10.5), risk ratio 1.27 (0.73–2.18); unadjusted two-proportion z-test p = 0.397.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a participant-level count: whether each child experienced a serious adverse event during the registered follow-up window. For each group, the registry reports the number of affected participants and the number included in the relevant analysis. A comparison then expresses the difference in observed event proportions between one group and the reference group as a risk difference. A risk difference is expressed on the probability scale; a negative value means the compared group had a lower observed proportion, while a positive value means it had a higher observed proportion.

What does the risk difference describe?

The risk difference is an absolute contrast between two proportions. It is distinct from a relative measure such as a risk ratio or odds ratio, which compares groups as a ratio. The posted comparisons include risk differences and, for most groups, relative measures as well. These alternative scales answer different descriptive questions, but neither scale removes limitations arising from differences in cohort membership, age group, or group size. The risk differences are based on the observed event counts and denominators in the relevant groups.

For example, the comparison of the older Cohort 1 RTS,S/AS02A group with the younger Cohort 1 RTS,S/AS02A reference has a risk difference of -0.086. Its confidence interval ranges from -0.137 to -0.036. The posted risk differences and intervals are statistical summaries of the between-group contrast; they do not show that the compared groups were otherwise interchangeable during the extension.

How should the p-values be read?

A p-value describes how compatible the observed data are with a specified null model of no difference, under the assumptions of the test used. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of a difference. The posted p-values are unadjusted. Some are reported as <0.001, while others include 0.065, 0.397, 0.004, 0.06, and 0.003. The registry record does not report a multiplicity procedure for these comparisons, so each value should be interpreted as a nominal result for its particular contrast, not as a family-wise assessment across all contrasts.

A p-value can be small when the estimate is precise enough relative to its uncertainty, and a larger p-value can occur when data are sparse or estimates are imprecise. In this trial, some groups contain relatively few participants or few recorded events, including groups with zero or one affected participant. For such comparisons, estimates and intervals can be sensitive to small changes in the observed counts. The uncertainty should be judged from the confidence intervals and group sizes, not from the p-value alone.

What does the confidence interval add?

A confidence interval gives a range of values compatible with the observed data under the statistical method used to construct it. It conveys uncertainty about the estimated contrast. A wide interval indicates less precision than a narrow interval, though the interval’s width also depends on the data and method. If an interval spans zero for a risk difference, it includes values representing no difference as well as values in either direction. That does not prove equality; it means the data do not pin down the contrast as precisely. The registry record does not describe the confidence-interval construction method.

What is meant by the planning illustrations?

The sample-size illustrations are calculations based on the observed effect, not the original trial’s design. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. The registry record does not report the original design assumptions or the original sample-size justification. These illustrations are hypothetical planning aids, not evidence that the study was designed around the displayed values.

General statistical point: over-dispersion matters when outcomes vary more than a model assumes, because a model that understates variability can make intervals too narrow and tests too optimistic. The primary outcome is a participant count with a binary event status, rather than a repeated-event rate for which over-dispersion would be the central issue. The registry record does not report an over-dispersion analysis. More generally, the choice of model and its assumptions affect uncertainty estimates; the available record does not give enough detail to evaluate the full analysis implementation.

The analysis population is described as the Total Vaccinated cohort with available data. This identifies who contributed to the primary analysis, but the registry record does not report the amount or pattern of missing data, whether any imputation was used, or whether sensitivity analyses assessed the effect of missing observations. Because the follow-up was open and took place after the earlier study, interpretation should also account for the fact that this endpoint reflects events over the extension period among participants who re-enrolled and had data available.

10. Confidence Intervals and Precision

The confidence intervals quantify uncertainty in the posted risk differences, in the units of an absolute difference in event proportions. They help show how much the data constrain each contrast. For Cohort 1 RTS,S/AS02A children aged 24 months or older compared with the reference, the interval runs from -0.137 to -0.036. For Cohort 1 Prevnar-Hiberix, the interval runs from -0.042 to 0.105. The first interval lies below zero, while the second includes zero and extends in both directions. This distinction describes the intervals; it does not establish causation or prove that two groups are equivalent.

Intervals for Cohort 2 comparisons include -0.162 to -0.029 for the younger RTS,S/AS02A group, -0.167 to -0.071 for the older RTS,S/AS02A group, -0.162 to -0.030 for the younger Prevnar-Hiberix group, and -0.146 to -0.035 for the older Engerix-B group. The Cohort 1 Engerix-B interval runs from -0.116 to -0.013. Group sizes and event counts vary substantially, so precision is not uniform across the comparisons. In particular, estimates based on groups with few participants or events warrant careful attention to their uncertainty.

A confidence interval is not a guarantee that the true value lies within its endpoints, and its width does not capture every source of uncertainty. The registry record does not report the method used to construct the intervals, or whether any adjustment was made for multiple comparisons. Nor do these intervals account for possible selection related to re-enrollment or missing data. They describe statistical uncertainty in the posted comparison under its analysis assumptions, rather than uncertainty about all aspects of study conduct or generalizability.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed contrast, not the original trial’s design. For these binary serious-event comparisons, the figures are participants per arm. They are not total event counts. The calculation assumptions stated for the illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The target effect is the stated fraction of the observed effect. The registry record does not report the original trial’s sample-size justification or original planning assumptions.

The 80% power column gives the estimated number of participants per arm that the planning calculation would require to have an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding planning number for a 90% chance. These are prospective design calculations for hypothetical studies, not a description of the probability that the completed study found a result. This is not observed power.

Rows that plan for 75% or 50% of the observed effect ask what would be needed to detect a smaller contrast than the one observed. Smaller effects are harder to distinguish from random variation, so the planning calculations generally require more participants. A curve, where displayed, is a visual way to follow how the required sample size changes as the assumed effect changes; it should be read in conjunction with the assumptions and the per-arm scale, not as a record of actual recruitment or a guarantee of future results.

The results are illustrations tied to the observed estimates, which can be imprecise, particularly in smaller groups. They do not demonstrate that the trial had adequate or inadequate power for its original aims. The record gives no original design assumptions against which to judge that question.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.119; treatment proportion p₁ = 0.033; average p̄ = 0.076.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3751 = 0.7353.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3701 = 0.3115.
4. n = (0.7353 + 0.3115)² / (-0.0863)² = 147.1, rounded up to 148 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.119318181818182                          # control proportion
p1 <- 0.0330097087378641                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(323622)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 148 (80% power), 197 (90% power)
 75% of observed effect: n per arm = 297 (80% power), 397 (90% power)
 50% of observed effect: n per arm = 743 (80% power), 994 (90% power)
power.prop.test: 147.1 
Simulated power at n = 148 per arm: 0.823 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple secondary outcomes in addition to the primary serious adverse event endpoint. They include antibody concentrations, time to malaria episodes under several case definitions, the number of clinical malaria episodes, and anemia. It also reports several comparisons for the primary safety endpoint, all using the same reference group. Testing multiple outcomes or contrasts increases the chance of observing an apparently notable result by chance if each is judged separately.

The registry record does not report a multiplicity adjustment, a hierarchy for testing outcomes, or a prespecified strategy for interpreting the secondary analyses. The posted p-values for the safety comparisons are unadjusted. They should therefore be read as individual nominal comparisons rather than as results that account for all outcomes and contrasts evaluated. The record also does not report whether any outcomes or analyses were selected after inspecting the data.

The primary analysis population was the Total Vaccinated cohort, defined as subjects vaccinated in the earlier study who re-enrolled for this follow-up and had data available. That description makes clear that the analysis did not include every enrolled child automatically: re-enrollment and data availability were part of the stated population definition. The registry record does not report the number of eligible children who did not re-enroll, the amount of missing data, reasons for missingness, or how missing observations were handled. It also does not report a sensitivity analysis comparing alternative missing-data assumptions.

These omissions limit assessment of the robustness of the estimates. If participants with available follow-up differed from those without it, the observed event counts might not describe all children from the earlier study equally well. The open follow-up and varying group sizes add context to that concern, but the record does not provide enough information to quantify any selection or missing-data effect. The statistical analysis plan and details of the testing procedure are not reported in the registry record.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: plasmodium falciparum infection (11.4%), anemia (8.8%), febrile convulsion (3.1%), pneumonia (1.9%), gastroenteritis (1.3%).

The serious adverse event results are participant counts over follow-up from Month 21 to Month 45. For example, the reference group had 21 affected participants among 176, while the Cohort 1 Engerix-B group had 29 among 528. The age-specific and cohort-specific group counts are listed separately; they should not be combined into an overall total without a defined analysis and denominator. A participant with an event is counted in the primary endpoint, which is not the same as a count of every event episode.

The registry also lists serious-event terms, including Plasmodium falciparum infection at 11.4%, anemia at 8.8%, febrile convulsion at 3.1%, pneumonia at 1.9%, gastroenteritis at 1.3%, and acquired immunodeficiency syndrome at 1.3%. These terms describe reported serious-event categories; they do not indicate that every event in a category was caused by a vaccine. The registry’s primary endpoint definition includes death, life-threatening occurrences, hospitalization or its prolongation, and disability or incapacity. The record does not supply additional detail on event attribution or individual event histories.

15. What the Result Does, and Does Not, Mean

The results describe serious adverse events recorded among children who re-enrolled in an open follow-up after an earlier vaccination study. They show how many participants in each registered group had at least one event and give statistical comparisons with the younger Cohort 1 RTS,S/AS02A group. The risk differences and confidence intervals summarize observed contrasts and their statistical uncertainty. They are not direct measures of vaccine efficacy, and the extension administered no new vaccine or drug.

Several comparisons have negative risk differences, while the Cohort 1 Prevnar-Hiberix comparison has a positive estimate and a confidence interval that includes zero. The posted estimates vary in precision, and some groups have few participants or events. These features matter when interpreting the apparent size and direction of differences. A small p-value does not establish clinical importance or causation, and a larger p-value does not establish that groups are equal.

The comparisons span different cohorts, age categories, and vaccine histories. The reference group is consistently the Cohort 1 RTS,S/AS02A group younger than 24 months, but the other groups do not all represent a single directly interchangeable comparator. The registry record does not report adjusted analyses that account for differences among groups, nor does it provide enough information about re-enrollment and missing data to assess possible selection effects. The open design also means masking was absent.

Accordingly, these findings should be understood as registry-posted safety follow-up data and unadjusted group contrasts. They do not by themselves show that a vaccine caused or prevented serious events, establish safety for populations beyond the enrolled children, or settle comparisons across the different groups. Such conclusions would require evidence and analysis beyond the results reported in this record.

16. Why This Trial Matters Statistically

Long-term follow-up raises a different statistical question from a short-term trial assessment. Here, the primary endpoint covers serious adverse events over a period from Month 21 to Month 45 after the first dose in the earlier study. Reporting participant counts for that interval lets readers see the size of the observed safety burden in each group and assess how much information supports each comparison.

The record also illustrates why group structure matters. The follow-up includes two cohorts, age strata, and different vaccine histories, while comparisons use one group as a common reference. Group sizes range from 42 to 528 participants in the posted arms, and event counts range from zero to 29. A contrast based on a small group with few events will generally be less stable than one supported by more observations; the confidence intervals make that uncertainty visible.

Statistically, the results also show why estimates, intervals, and p-values should be read together. A risk difference describes an absolute contrast, its confidence interval conveys precision, and a p-value reflects compatibility with a null model under a test’s assumptions. None alone answers whether a difference is clinically important or causal. Multiple comparisons and unreported analysis details further constrain interpretation.

The study is relevant as an example of the information a registry can preserve about follow-up: the endpoint definition, group counts, analysis population, and some uncertainty estimates. At the same time, the record leaves important questions about missing data, multiplicity, and the original statistical plan unanswered. Those gaps do not erase the descriptive value of the results, but they delimit the conclusions that can be drawn from them.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The extension was open and unmasked, so knowledge of group history was not concealed during follow-up. The groups also differ by cohort, age at first vaccination, and prior vaccine history. Although the study is described as randomized, the registry record does not provide enough detail about the allocation procedure or the relation between original assignments and extension analyses to assess how well the posted contrasts preserve the benefits of randomization.

Re-enrollment and availability of data define the primary analysis population. The record does not state how many children did not re-enroll or why, how much data were missing, or whether missingness differed among groups. Such differences could affect the representativeness of the observed counts. The registry record does not report a missing-data sensitivity analysis.

Several groups have small denominators and very few events. Estimates from these groups are correspondingly uncertain, as reflected in their confidence intervals. The record does not describe the interval construction method or the full statistical model. It also does not report a multiplicity correction despite multiple outcomes and comparisons, or a complete prespecified statistical analysis plan.

The primary endpoint counts participants with serious adverse events and does not provide a complete account of event timing, recurrence, severity beyond the registry definition, or causal attribution. The serious-event terms are listed, but detailed case-level information is not reported in the registry record. The follow-up did not administer new vaccines, so the results describe events during later observation after the earlier study’s vaccinations rather than the safety of a new dosing period in this extension.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Long-Term Follow-up of Children for a 2-Year Period to Confirm the Safety and Immunogenicity of GSK 257049 Vaccine — NCT00323622.
• Registry record last updated 2016-12-09; status last verified 2016-10.
• Results first posted 2014-03-27.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT00323622 is a completed, randomized, open, parallel phase II follow-up study sponsored by GlaxoSmithKline. It enrolled 1,737 participants and followed children in Mozambique from Month 21 to Month 45 after Dose 1 in the earlier NCT00197041 study. No new subjects were recruited and no vaccine was administered during the extension.

The primary outcome was the number of subjects with serious adverse events, analyzed in the Total Vaccinated cohort with available data. The registry reports counts by cohort, age group, and prior vaccine group, plus risk-difference comparisons against the Cohort 1 RTS,S/AS02A group younger than 24 months. The comparisons are unadjusted, and the record does not report the original sample-size justification, multiplicity strategy, or missing-data sensitivity analyses.